Interaction between Transcription Factor, Basal Transcription Factor 3, and the NH2-Terminal Domain of Human Estrogen Receptor A

The estrogen receptor (ER), like other members of the nuclear receptor superfamily, possesses two separate transcriptional activation functions, AF-1 and AF-2. Although a variety of coactivators and corepressors of AF-2 have been identified, less is known of the mechanism of action of AF-1. We have used the yeast two-hybrid system to isolate a cDNA coding for a protein that binds specifically to the AF-1 region of human ERA. This cDNA codes for the transcription factor basal transcription factor 3 (BTF3). The specificity of the interaction between BTF3 and ERA has been confirmed in vivo and in vitro . Transient transfection experiments reveal that overexpression of BTF3 modulates the transcriptional response of reporter genes to ERA. BTF3 interacts with ERA that has been activated either by 17B-estradiol (ligand-dependent activation) or by epidermal growth factor (ligand-independent activation). The effects of BTF3 on the reporter genes requires the presence of ERA containing an active AF-1 function. BTF3 may be a component of the mechanism by which the AF-1 function of ERA stimulates gene transcription. (Mol Cancer Res 2007;5(11):1191–200) Introduction Steroid hormones, such as 17h-estradiol, produce responses in their target cells by regulating gene expression (1, 2). The receptors for steroid hormones belong to a superfamily of nuclear receptors that, upon binding their individual ligand, are able to bind to specific DNA sequences, hormone response elements, and stimulate the initiation of transcription of nearby hormone-responsive genes (3, 4). Members of the receptor superfamily contain two separate regions that are involved in the activation of transcription (4, 5). These two transactivation functions (AF-1 and AF-2), which are physically distinct from each other, show differences in their mode of action. In particular, the activity of AF-2 is dependent on ligand binding by the receptor, whereas AF-1 activity is ligand-independent (6, 7). More recently, it has been shown that non–ligand signal molecules such as certain growth factors (8-14) may regulate AF-1 activity. The mechanism by which the receptor, bound to its response element, may influence the assembly/stability of the multiprotein transcription initiation complex, situated at the gene promoter, is unclear. Transcriptional interference (‘‘squelching’’) experiments have provided evidence that both AF-1 and AF-2 require coregulator molecules to achieve stimulation of transcription (15, 16). Subsequently, a variety of molecular candidates for AF-2 coregulators, both coactivators and cosuppressors, have been described (17-21). Such coregulator proteins bind in a ligand-dependent manner to regions of the amino acid sequence that are conserved between receptors. Various mechanisms have been suggested by which they may contribute to transcriptional activation (22, 23). The mechanism of action of the AF-1 transactivation function is more obscure, not least because of the lack of sequence conservation at the amino-terminal end of nuclear receptors, where AF-1 is located. Fewer coregulator proteins for AF-1 have been described (11-14), although it has recently been reported that the steroid receptor RNA activator molecule acts as an AF-1 coactivator (21), and that certain AF-2 coactivators may also bind to the A/B domain of estrogen receptor a (ERa; refs. 24, 25). ER AF-1 binds the COOH terminus of glucocorticoid receptor– interacting protein 1 (NID/AF-1) and other p160s (24). ERa AF-1 transcriptional activity is also enhanced by p300 and DEAD box protein p68/p72, which form a protein complex with p160 family proteins and p300/CBP, and directly bind to the A/B domain to potentiate AF-1 activity (11, 26). The phosphorylation of the serine residue at position 118 in the A/B domain stabilizes the complex formation of ER and the coactivator complex containing p68/p72 to potentiate the AF-1 activity (19, 26). Phosphorylation of certain serine residues, within the A/B domain of several nuclear receptors, may play an essential role in the action of AF-1 (10, 26, 27). It has been shown that a point mutation that replaces Ser of human ERa with an unphosphorylatable Ala residue prevents the activation of the receptor by epidermal growth factor (EGF; refs. 10, 26). We have therefore used the yeast two-hybrid system (28) to isolate cDNA coding for a protein that binds specifically to the region of the human ER that contains the AF-1 transactivating Received 3/12/07; revised 6/12/07; accepted 7/5/07. Grant support: North West Cancer Research Fund, RRG, NI and Action Cancer, Northern Ireland. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Requests for reprints: Mohamed K. El-Tanani, Centre for Cancer Research and Cell Biology, Queen’s University Belfast, University Floor, Belfast City Hospital, Lisburn Road, Belfast BT9 7AB, United Kingdom. Phone: 44-28-9026-3486; Fax: 44-9026-3744. E-mail: [email protected] Copyright D 2007 American Association for Cancer Research. doi:10.1158/1541-7786.MCR-07-0123 Mol Cancer Res 2007;5(11). November 2007 1191 on June 20, 2017. © 2007 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from function. This cDNA encodes a previously identified transcription factor which is a 27 kDa protein, the precise function of which is unclear (29). In human cells, alternative splicing of the initial transcript of the basal transcription factor 3 (BTF3) gene seems to give rise to mRNAs coding for two versions of the protein, one (BTF3a) encoding the full-length protein (206 amino acids), the other (BTF3b) lacking the first 44 amino acids (29). Our cDNA represents a previously reported (29) version of BTF3b mRNA that possesses a different 5¶ untranslated sequence from that originally described. Transient transfection experiments showed the functional significance of the association of BTF3 with the ER. Results Isolation of cDNA Coding for a Protein that Binds to the A/B Domain of the ERa A commercial yeast two-hybrid system (see Experimental Procedures) was used to identify human cDNA coding for proteins that bind to the A/B domain (containing the AF-1 function) of the human ERa. A cDNA library constructed from EGF-stimulated MCF-7 human breast cancer cells was inserted into the ‘‘target’’ vector whereas cDNA coding for amino acids 1 to 185 (the complete A/B domain) of human ERa was inserted into the ‘‘bait’’ vector. The two recombinant vectors were coexpressed in yeast cells and f10 yeast transformants were screened. Four hundred yeast cells were selected on the basis of their ability to grow in medium lacking histidine, i.e., demonstrating activation of the HIS3 reporter gene. These were further screened for their ability to express a second reporter gene, lacZ, and six positive clones were identified. Characterization of the Isolated cDNA The DNA sequences of the inserts in three of the clones (f1, f2, and b3) were determined and found to contain the same sequence (875 bp in length). A BLAST search of the GenBank nucleotide database revealed that this sequence was 99% identical to a previously reported variant of the BTF3b transcription factor (29). When clone f1 DNA was hybridized to a Northern blot of total RNA from the estrogen-responsive human breast cancer cell line, MCF-7, it produced a single band of the expected size (f0.9 kb) for BTF3 mRNA (Fig. 1). Control (1-fold induction) and estradiol (4.2-fold induction) cells were tested. High expression of BTF3 was also observed in cells treated with EGF (3.9-fold induction). The specificity of the binding of BTF3b to the A/B domain of ERa was determined by returning the f1 plasmid to yeast cells in the presence of a series of bait vectors containing cDNAs coding for alternative protein fragments (Fig. 2A). Expression of the lacZ reporter gene occurred only in yeast cells containing the BTF3 plasmid in the simultaneous presence of an A/B domaincontaining bait plasmid. Significantly, although binding was evident between BTF3b and full-length ERa, no binding was seen with receptor lacking the A/B domain (domains C/D/E/F, amino acids 179-595). These data indicate that BTF3b binds specifically to the region of the ER containing the AF-1 function. We confirmed and extended these results by examining the specificity of BTF3b binding using an alternative version of the two-hybrid system in mammalian cells. Figure 2B presents the results of BTF3b-binding experiments in MCF-7 cells. Similar results were obtained with HeLa cells (data not presented). Activation of the reporter gene was only detected when BTF3 was coexpressed with the A/B domain of ERa. Interestingly, no binding could be detected between BTF3b and the A/B domain of ERh. In a pilot study, we have shown no binding between BTF3 and vitamin D receptor (VDR), retinoid X receptor, peroxisome proliferator-activated receptor, glucocorticoid receptor (GR), and androgen receptor (AR) (data not shown). It should be noted that when transfected without BTF3, the A/B domain of ERa, but not of ERh, caused a small but consistent increase in reporter gene activity. Binding between BTF3b and the A/B domain of ERa was also shown in vitro (Fig. 3A). MCF-7 cells were separately transfected with either the BTF3b-containing target plasmid or a target plasmid containing the BTF3b sequence inserted in the antisense orientation or an expression plasmid encoding fulllength ERa. Whole-cell extracts prepared from the transfected MCF-7 cells were mixed and incubated in the presence of an antibody against BTF3 (Santa Cruz Biotechnology, Inc.). The immunoprecipitated proteins were separated by PAGE and subjected to Western blotting using a monoclonal antibody against human ERa (Fig. 3A, top). Immunoprecipitate was analyzed by Western blot using BTF3 antibody as a control (Fig. 3A, bottom). This data clearly shows that ERa is specifically coimmunoprecipitated with BTF3b. In this experiment, we did not observe coimmunoprecipitation of the endogenous ERa from MCF-7 cells when using the BTF3 antibody. This is probably due to the association of unliganded ERa with HSPs and the requirement for estradiol to dissociation of this complex and stabilization of the receptor. In agreement with this, we showed that the binding in vivo was enhanced by the ligand (Fig. 3C). To investigate whether BTF3a and ERa could interact directly, coimmunoprecipitation experiments were done in cell-free protein-synthesizing reticulocyte lysates. Control lysates contained either ERa (Fig. 3B, lane 1) or S-BTF3 (Fig. 3B, lane 2) and produced proteins of the correct size. FIGURE 1. Specificity of BTF3 binding to the ER. Estradiol and EGF induction of BTF3 mRNA in MCF-7 cells. MCF-7 cells were cultured in estrogen-free medium for 6 days (control) and were then transferred to medium supplemented with estradiol (E2, 10 8 mol/L) or EGF (100 ng/mL), for a further 24 h. Total RNA was isolated from the cells and subjected to Northern blot analysis. After hybridization with probes for BTF3 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH ), each autoradiograph was quantified by densitometry. BTF3 results were normalized relative to the glyceraldehyde-3-phosphate dehydrogenase signals and are expressed as fold increase above the control cell result. A representative Northern blot illustrating the induction of BTF3 mRNAs by estradiol (E) and EGF compared with mRNA levels in untreated cells (C ). Green et al. Mol Cancer Res 2007;5(11). November 2007 1192 on June 20, 2017. © 2007 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from S-BTF3a–containing lysates were incubated with nonradioactive ERa-containing or unprogrammed lysates. The resultant protein interaction complexes were immunoprecipitated with anti-ERa antibody (30) and analyzed on polyacrylamide gels (Fig. 3B). An antibody to ERa coimmunoprecipitated a S-labeled protein of 27 kDa from BTF3a-producing lysates (Fig. 3B, lane 3). These S-27 kDa proteins corresponded to the major radioactive protein in the lysates and possessed the same molecular weight as reported for human BTF3a (31). The specificity of the immunoprecipitated complexes formed between BTF3a and ERa was tested as follows. When Slabeled BTF3a-containing lysates were incubated with unprogrammed nonradioactive lysates and then immunoprecipitated with anti-ERa, no radioactive BTF3a protein of 27 kDa was observed on the resultant polyacrylamide gels (Fig. 3B, lane 4). This radioactive band was not observed in controls in which an unrelated transcription factor Tip 60-HA (ref. 32; Fig. 3B, lane 5) or unprogrammed lysates—hemagglutinin (Fig. 3B, lane 6) replaced those containing ERa. Immunoprecipitate was analyzed by Western blot using ERa and hemagglutinin antibodies as controls (Fig. 3B, bottom). To determine whether interaction occurred between endogenous ERa and endogenous BTF3 in untransfected MCF-7 cells, cell lysates were immunoprecipitated with control IgG or ERa antibody, followed by Western blotting of the precipitates with antibodies to ERa and to BTF3 as indicated. A strong FIGURE 2. A. Yeast two-hybrid system. Yeast cells were cotransfected with combinations of bait and target plasmids and the activation of the h-galactosidase reporter gene was determined by quantitative enzyme assay. Results are expressed as a percentage of the positive control activity (p53 + SV40). All combinations are indicated as (bait plasmid) + (target plasmid). Bait plasmids: A/B, A/B domain of ERa; ER, full-length ERa; CDEF, domains C, D, E, and F of ERa; Lamin, human lamin C; p53, murine p53. Target plasmids: BTF3, full-length BTF3b; SV40, SV40 large T-antigen. 0, absence of second plasmid. *, P < 0.0001, significant difference from the A/B alone result. B. Mammalian two-hybrid system. MCF-7 cells were cotransfected with combinations of bait and target plasmids together with the reporter plasmid. The activity of the reporter gene was determined by quantitative CAT assay. Results are presented as a percentage of the positive control activity (p53 + pVP16T). Bait plasmids: A/B(a), A/B domain of ERa; ER(a), full-length ERa; CDEF(a), domains C, D, E, and F of ERa; A/B(b), A/B domain of ERh; ER(b), full-length ERh; p53, murine p53; pM, bait plasmid without insert. Target plasmids: BTF3, full-length BTF3b; pVP16, target plasmid without insert; pVP16T, SV40 large T-antigen. *, P < 0.001, significant difference from the A/B(a) alone result. BTF3 Interacts with Human Estrogen Receptor a Mol Cancer Res 2007;5(11). November 2007 1193 on June 20, 2017. © 2007 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from interaction between endogenous ERa and BTF3 was observed in both estradiol-treated and untreated MCF-7 samples (Fig. 3C). This interaction was dependent on the presence of ERa in the complex, without it, no interaction occurred (Fig. 3C). The association between ERa and BTF3 was significantly enhanced by the presence of estradiol, whereas the presence of an antiestrogen, such as ICI 164384 was observed to abrogate this interaction (data not shown). These results confirm that endogenous ERa and BTF3 could form interacting complexes in untransfected MCF-7 cells. Modulation of ER Function by BTF3 The functional significance of the binding of BTF3 to the ER was investigated in transient transfection experiments. The cDNA encoding BTF3b or the active form BTF3a were cloned into the mammalian expression vector pBK-CMV. We cotransfected MCF-7 cells with ERE.VIT.CAT, an estrogeninducible reporter construct (33) and HEG0, which encodes the wild-type human ERa (34), in the presence of a range of amounts of the BTF3b or BTF3a expression vector. CAT activity was measured after 48 h of exposure of the transfected cells to estradiol or EGF (Fig. 4A and B). The overexpression of BTF3b (Fig. 4A) or BTF3a (Fig. 4B) in MCF-7 cells increased the stimulation of reporter gene transcription by both estradiol and EGF in a dose-dependent manner. Cells transfected with 1 ng/30 mm dish (f1.1 10 cells) of the BTF3b or BTF3a plasmid showed 2and 3.2-fold increases (P = 0.0003 and P = 0.0007, respectively) in estrogen-induced CAT activity and 3and 7.5-fold increases (P = 0.0008 and P = 0.0002, respectively) in FIGURE 4. The effect of BTF3b (A) or BTF3a (B) overexpression on reporter gene induction in MCF-7 cells. MCF-7 cells were depleted of steroid hormones, transferred to serum-free medium, and then cotransfected with ERE.VIT.CAT, HEG0, and the indicated amounts of BTF3b or BTF3a DNAs. Transfected cells were cultured for 24 h in medium containing 10 nmol/L of estradiol (n) or 100 ng/mL of EGF (5), or without any addition (o). Cells were then harvested and assayed for CAT and h-galactosidase activity. CAT activity is expressed relative to the activity (100%) in cells treated with estradiol without BTF3. FIGURE 3. A. Immunoprecipitation and Western blot. Detection of ERa and BTF3b protein lysates by its respective antibody in immunoblots (lanes 1 and 2), respectively. Whole cell extracts from MCF-7 cells expressing either full-length ERa or BTF3 or ‘‘reversed’’ BTF3 were mixed and subjected to immunoprecipitation with an antibody against BTF3. Immunoprecipitated proteins were analyzed by Western blot using an antibody against ERa. Lane 3, ERa extract + BTF3 extract. Lane 4, ERa extract + ‘‘reversed’’ BTF3 extract. Lane 5, BTF3 extract alone. The positions of protein molecular weight markers are indicated. B. Coimmunoprecipitation of ERa with BTF3a. BTF3a S-labeled and nonradioactive ERa protein were produced in cell-free, protein-synthesizing reticulocyte lysates. Detection of nonradioactive ERa protein lysate by its respective antibody in immunoblots (lane 1) or BTF3a S-labeled lysate (lane 2). Coimmunoprecipitation of lysates containing expression vectors for S-labeled BTF3a and nonradioactive ERa (lane 3). Coimmunoprecipitation of lysates containing expression vectors for S-labeled BTF3a and unprogrammed lysates (lane 4 ), nonradioactive Tip 60-HA (lane 5), and unprogrammed lysates-hemagglutinin (lane 6 ). Immunoprecipitation was by monoclonal antibody to ERa (lanes 3 and 4), a polyclonal antibody to hemagglutinin (Santa Cruz Biotechnology; sc-805; lanes 5 and 6 ), and the resultant precipitates run on polyacrylamide gels. The position of S-labeled BTF3a protein is indicated (arrow) and its molecular weight shown. C. Coimmunoprecipitation of BTF3 with ERa in MCF-7 cells. Total cell lysates were immunoprecipitated with either control IgG or BTF3 antibody. The immunoprecipitates were resolved by SDS-PAGE and analyzed by Western blotting using anti-BTF3 or anti-ERa antibodies as indicated. IP, immunoprecipitation; WB, Western blotting. Green et al. Mol Cancer Res 2007;5(11). November 2007 1194 on June 20, 2017. © 2007 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from EGF-induced CAT activity. Transfection with larger amounts of BTF3b or BTF3a vector reduced the level of CAT induction. It would therefore seem that the stimulation of transcription by BTF3b or BTF3a requires the presence of ERa that has been activated, by either a ligand-dependent or a ligand-independent mechanism, and is not simply a consequence of an increase in efficiency of the basic transcription machinery. The Receptor Domain Specificity of the BTF3 Effect on ER Activity in MCF-7 Cells Transient transfection studies, using two deletion mutants and a point mutant of ERa, were carried out to establish which transactivation functions were affected by the presence of overexpressed BTF3b or BTF3a. HE15 contains an insert coding for a truncated ERa (amino acids 1-282), which lacks the ligand-binding domain and AF-2 but retains AF-1 (6). HEG19 codes for a truncated ERa (amino acids 179-595), which lacks AF-1 but retains ligand-binding and AF-2 (4). HE457 codes for a full-length ER containing a point mutation that substitutes Ala for Ser at residue 118 within the A/B domain (27). MCF-7 cells were cotransfected with the ERE.VIT.CAT reporter construct, 1 ng of BTF3b (Fig. 5A) or BTF3a (Fig. 5B) plasmid and either wild-type ERa (HEG0) or mutant ERa plasmid, and were then cultured for 24 h in the presence or absence of estradiol (Fig. 5A and B). BTF3b (Fig. 5A) or BTF3a (Fig. 5B), in the presence of wild-type ERa, increased the response of ERE.VIT.CAT to estradiol. In the presence of ERa lacking ligand-binding activity, no response to estradiol was seen and BTF3b or BTF3a were without effect. A receptor lacking AF-1 was able to support a low level of induction by estradiol but was unaffected by BTF3b or BTF3a overexpression. The response of the reporter gene to estradiol, in the presence of HE457 was unaffected by BTF3b or BTF3a overexpression. Apparently, this point mutation in the A/B domain significantly reduces the sensitivity of the ER to BTF3b and BTF3a when activated by estradiol. These results were compared with those from similar experiments in which the transfected MCF-7 cells were incubated with EGF to achieve ligand-independent activation of ERa (Fig. 5C and D). In the presence of the wild-type receptor (HEG0), EGF achieved a significant induction (P = 0.003 and P = 0.002, respectively) of ERE.VIT.CAT. This response to EGF was modified by BTF3b (Fig. 5C) and BTF3a (Fig. 5D) overexpression in the same manner as the response to estradiol, i.e., BTF3b or BTF3a increased the induction of ERE.VIT.CAT by EGF. EGF activated the ERa primarily by its effect on AF-1 activity (9, 10) and was therefore able to induce reporter gene activity in the presence of the truncated receptor HE15. Overexpression of BTF3b and BTF3a had a positive effect on EGF induction of ERE.VIT.CAT in the presence of HE15. HEG19 encodes a receptor lacking AF-1, which fails to support reporter gene induction by EGF, showed no response to BTF3b or BTF3a overexpression. The point mutation in HE457, which has been reported previously (10, 26), abolishes ligand-independent activation of the ER by EGF and overexpression of BTF3b or BTF3a failed to restore EGF responsiveness. These results confirm that the effects of BTF3b and BTF3a on reporter gene expression are exerted via AF-1 and indicate that AF-1 needs to be ‘‘activated’’ if it is to respond to BTF3 overexpression. Discussion The A/B domain has been shown to contain transactivating activity (AF-1) in most, if not all, members of the nuclear receptor superfamily (3, 6). However, it is also the region of these receptors that shows least conservation in length or sequence (35). The E domain, which contains the second transactivating function (AF-2) as well as ligand-binding activity, is more tightly conserved and an amphipathic a-helix structure has been identified within the domain that seems to be essential for AF-2 activity (36). Evidence has been presented that AF-1 and AF-2 interact to produce the full transactivating function of the wild-type ER (5, 6, 37), although deletion mutants of the receptor indicate that AF-1 and AF-2 could act independently of each other. AF-2 stimulation of the initiation of transcription is believed to occur by its ligand-dependent recruitment of a multiprotein coactivator complex and a variety of AF-2 coactivators (and corepressors) have been identified (17-21). Transcriptional interference (squelching) experiments also indicate that AF-1 acts via coregulators (11-16) but fewer candidates have been identified. We used the yeast two-hybrid system (28) to isolate a cDNA coding for a protein that binds specifically to the region of the human ER containing the AF-1 transactivation function. Determination of the sequence of the cDNA revealed that it originated from mRNA for the BTF3b transcription factor (29). No binding of this protein to the remainder of the ER could be detected by the yeast two-hybrid system. This was confirmed using the two-hybrid system in mammalian cells. ERa and ERh have considerable divergence of sequence in their A/B domains and it is therefore interesting that no binding between this region of ERh and BTF3 was detected. Binding of BTF3 to the ER seems to occur in a sequence-specific manner. BTF3 is a 27 kDa protein whose precise function is unclear. It was originally identified as a basal transcription factor required for the accurate initiation of transcription by RNA polymerase II (31). However, subsequent studies indicated that BTF3 was not in fact essential for specific, in vitro initiation of transcription (38). Its biological importance is shown by the fact that mouse embryos, homozygous for a deletion of the BTF3 gene, die at an early stage after implantation (39). BTF3 is evolutionarily conserved and is present in a variety of mammalian cells (29). In yeast (Saccharomyces cerevisiae) there are two homologous genes, although yeast BTF3 inhibits the transcription of specific genes in yeast cells (40). In human cells, alternative splicing of the initial transcript of the BTF3 gene seems to give rise to mRNAs coding for two versions of the protein, one (BTF3a) encoding the full-length protein (206 amino acids), the other (BTF3b) lacking the first 44 amino acids (29). Our cDNA represents a previously reported (29) version of BTF3b mRNA that possesses a different 5¶ untranslated sequence from that originally described. We investigated the functional role of BTF3b and BTF3a in ER activation of gene transcription by overexpressing the genes in cells cotransfected with ER and estrogen-responsive reporter BTF3 Interacts with Human Estrogen Receptor a Mol Cancer Res 2007;5(11). November 2007 1195 on June 20, 2017. © 2007 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from genes. This revealed that BTF3 (b and a) could exert a positive, dose-dependent effect on reporter gene transcription only in the presence of activated ERa. When considering such results, it should be borne in mind that MCF-7 cells express BTF3 (b and a) themselves (29), so that any effect of overexpression of the transfected gene is superimposed on the action of the endogenous level of BTF3 protein(s). At higher levels of transfected BTF3b and BTF3a (>1.0 ng/10 cells), the increase FIGURE 5. ER dependence of the effect of BTF3 on reporter gene response to estradiol or EGF in MCf-7 cells. MCF-7 cells were depleted of steroids and then transfected with ERE.VIT.CAT reporter gene, with an ER expression vector as indicated: (n) HEG0, (5) HE15, ( ) HEG19, or ( ) HE457, with and without BTF3b (A) or BTF3a (B) DNA (1 ng/30 mm dish). Transfected cells were cultured for 24 h in medium with 10 nmol/L of estradiol (E) or without addition (C). CAT activity in the harvested cells was assayed and is expressed relative to the activity in cells cultured with estradiol after transfection with wild-type ER (HEG0 ) without BTF3b or BTF3a. *, P < 0.001, significant difference from the estradiol result for cells not transfected with BTF3. MCF-7 cells were depleted of steroids and then transfected with ERE.VIT.CAT reporter gene, with an ER expression vector as indicated: (n) HEG0, (5) HE15, ( ) HEG19, or ( ) HE457, with and without BTF3b (C) or BTF3 (D) DNA (1 ng/30 mm dish). Transfected cells were cultured for 24 h in medium with 100 ng/mL of EGF (EGF ) or without addition (C ). CAT activity in the harvested cells was assayed and is expressed relative to the activity in cells cultured with estradiol after transfection with wild-type ER (HEG0 ) without BTF3b or BTF3a. *, P < 0.001, significant difference from the EGF result for cells not transfected with BTF3. Green et al. Mol Cancer Res 2007;5(11). November 2007 1196 on June 20, 2017. © 2007 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from in estradiol-induced transcription of the ERE.VIT.CAT reporter declines, possibly indicating that the excess BTF3 protein is in some way competing with the ER-bound protein for functional targets. These data suggest that the ER was squelched by high levels of BTF3 and its ability to sequester target proteins required for transcriptional activation by the receptor (41). We observed no inhibitory effect of BTF3 overexpression on MCF-7 cell growth (data not shown). The use of deletion mutants of the ERa confirmed that the BTF3 effect on reporter gene transcription requires the presence of the A/B domain (AF-1) of the ER. In addition, the response to BTF3 (b and a) also requires that the AF-1 exists in an ‘‘activated’’ state, either as the result of ligandbinding (HEG0 + estradiol) or as the result of ligandindependent activation (HE15 + EGF). The point mutation (HE457) that converts Ser to an unphosphorylatable Ala reduces but does not prevent the response of estradiolactivated ER to BTF3 overexpression (compare Fig. 5A and B and Fig. 5C and D). However, this mutation completely eliminates the response of EGF-activated ER to BTF3 (compare Fig. 5A and B and Fig. 5C and D). It seems that phosphorylation at Ser may be important for productive interaction with BTF3 but that estrogen-binding promotes interaction in an alternative manner. In summary, we present evidence suggesting the existence of a mechanism by which the AF-1 transactivation function of the human ER may regulate gene transcription. This mechanism involves the BTF3 ‘‘basal’’ transcription factor in a new role. Experimental Procedures Chemicals and Materials Tissue culture medium, newborn calf serum, FCS, antibiotics, and trypsin-EDTA were purchased from Life Technologies. 17h-Estradiol and EGF (human, recombinant) were purchased from Sigma. Restriction enzymes, DNA-modifying enzymes, and Taq polymerase were obtained from Boehringer Mannheim. Expression Vectors and Reporter Constructs The ERa plasmids and estrogen-responsive reporter constructs were obtained from the laboratories in which they were originally constructed. HEG0 (34) encodes a full-length (amino acids 1-595), wild-type human ERa. HE15 (5) encodes amino acids 1 to 282 and HEG19 (34) encodes amino acids 179 to 595 of the human ERa. HE457 (27) encodes a fulllength human ERa containing a point mutation that replaces Ser by an alanine residue. ERE.VIT.CAT contains a promoter that is derived from the Xenopus laevis vitellogenin B1 gene (33). Yeast hybrid-protein expression vectors, pBDGAL4 and pADGAL4, were obtained from Stratagene as components of the HybriZAP kit. The pBK-CMV mammalian expression vector was obtained from Stratagene and the pSV-hgalactosidase plasmid was obtained from Promega. Mammalian hybrid-protein expression vectors, pM and pVP16, together with the reporter gene vector, pG5CAT, were obtained as part of the Mammalian MATCHMAKER kit (Clontech Laboratories). Construction and Screening of the cDNA Library The HybriZAP two-hybrid system (Stratagene) was used, according to the manufacturer’s instructions, to screen a cDNA library constructed from polyA RNA isolated from MCF-7 cells. The cells had been depleted of steroids, transferred to serum-free medium, and then cultured for 24 h in the presence of EGF, as previously described (10). The cDNA was directionally inserted into EcoRI/XhoI-digested pADGAL4 vector. Human ERa A/B domain DNA (coding for amino acids 1-185) was prepared by PCR amplification from plasmid HEG0 DNA. The A/B DNAwas inserted into the EcoRI site of the pBDGAL4 vector and the correct orientation and insertion of the DNA was confirmed by determining the nucleotide sequence of the recombinant vector across the EcoRI site. After cotransformation of yeast YRG-2 cells with the two vectors, cells were selected for growth in medium lacking histidine (activation of HIS-3 reporter gene, indicating cDNA-encoded protein binding to the hybrid ‘‘bait’’ protein). From f10 transformed yeast cells, 400 cells were selected in this manner. These were then assayed for their expression of the hgalactosidase reporter gene by filter lift assay—the production of blue coloration with X-gal substrate indicating a positive result. Six yeast clones expressing h-galactosidase were identified. Isolation and Sequencing of cDNA Clones For the cloning of BTF3a, total RNA was isolated from HeLa cells using the guanidine isothiocyanate/acid phenol method (42) and used to synthesize the cDNA using avian myeloblastosis virus reverse transcriptase (43). An aliquot of cDNA was amplified by PCR using the primers oligonucleotide, forward: 5¶ATGCGACGGACAGGCGCA3¶ and reverse: 5¶TCAGTTTGCCTCATTCTT3¶, derived from BTF3a genes. The PCR incubations were carried out for 36 cycles of denaturation (94jC for 1 min), annealing (46jC for 1 min), and elongation (72jC for 2 min) using a reaction mixture containing 500 pmol of each oligonucleotide, 200 Amol/L of deoxynucleotide triphosphate, and a reaction buffer provided with the Taq DNA polymerase (Hoffmann-La Roche). PCR products were isolated from agarose gel and cloned directly into the pBKCMV expression vector (Stratagene) following the manufacturer’s instructions. After transformation, a limited number of positive clones were selected and sequenced to identify the PCR products corresponding to BTF3a fragments. Determination of Binding Specificity In vivo The p53 (DNA coding for amino acids 72-390 of murine p53 within pBDGAL4 vector), pLamin C (DNA coding for amino acids 67-230 of human lamin C within pBDGAL4), and pSV40 (DNA coding for amino acids 84-708 of the SV40 large T-antigen within pADGAL4 vector) control plasmids were obtained from Stratagene as components of the HybriZAP yeast two-hybrid kit. The ER coding DNAwas released from HEG19 plasmid by EcoRI digestion and then ligated with pBDGAL4 vector to create the ‘‘domains C/D/E/F’’ bait plasmid. The quantitative liquid h-galactosidase assay was conducted according to protocol 10 of Olesen et al. (44). The positive control plasmids, pM-53 (a fusion of GAL4 DNA binding domain to murine p53) and pVP16-T (a fusion of BTF3 Interacts with Human Estrogen Receptor a Mol Cancer Res 2007;5(11). November 2007 1197 on June 20, 2017. © 2007 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from GAL4 activating domain to SV40 large T antigen, which interacts with p53) were components of the Mammalian MATCHMAKER two-hybrid kit. DNA coding for the complete A/B domain (amino acids 1-185) of ERa was amplified by PCR of HEG0 using primers that incorporated EcoRI (5¶ primer) or SalI (3¶ primer) sites. The amplified product was digested with EcoRI and SalI and ligated to plasmid pM, digested with the same restriction enzymes. In the same manner, the sequence coding for full-length BTF3b was amplified by PCR of plasmid pAD-GAL4 f1 and inserted into EcoRI/SalI digested pVP16. The complete coding sequence of HEG19 was released by digestion with EcoRI and SalI and then ligated to the pM vector, digested with the same enzymes. The coding sequence for amino acids 1 to 90 of ERh was amplified by PCR of plasmid hERh0 (45) and inserted into pM vector, digested with EcoRI and SalI. MCF-7 cells were cotransfected with combinations of bait and target plasmids, the reporter plasmid and pSV-h-galactosidase plasmid (Promega) as previously described (2, 10). Cells were cultured for 48 h and then harvested and CAT (liquid scintillation method) and hgalactosidase activities were assayed in whole cell extracts using assay kits (Promega) according to the manufacturer’s protocols. CAT activity results were normalized relative to the h-galactosidase activity and are expressed as a percentage of the positive control level (pM-53 + pVP16T). Results are the mean F SD of at least three separate experiments. Immunoprecipitation and Western Blot for BTFb and ERa MCF-7 cells were transfected either with an expression plasmid encoding full-length ERa or with the f1 target plasmid encoding BTF3b or with the target plasmid containing the BTF3b cDNA inserted in an antisense orientation (‘‘reversed’’). After incubation for 48 h, cells were harvested and whole cell lysates were prepared according to the instructions for immunoprecipitation provided by Santa Cruz Biotechnology. Extracts were mixed and incubated with antibody against BTF3 (Santa Cruz Biotechnology). Immunoprecipitation and Western blot analysis, using antibody against ERa, were carried out according to the instructions of the supplier of the anti-ERa antibody (Santa Cruz Biotechnology). Endogenous Immunoprecipitation and Western Blotting MCF-7 cells were treated with or without 10 8 mol/L of estradiol, harvested, and lysed in immunoprecipitation buffer [0.5% (v/v) NP40 solution: 50 mmol/L Tris-HCl, 150 mmol/L NaCl, 1 mmol/L EDTA, 0.5% (v/v) NP40, 10% (v/v) glycerol, and protease inhibitors]. Total cell lysates were incubated with protein G beads with normal IgG as a pretreatment. The supernatants were then treated with ERa antibody and immunoprecipitated with protein G beads for 3 h. After immunoprecipitation, beads were washed four times. Laemmli buffer was added, and the samples were loaded onto 6% to 10% SDS-PAGE gel and transferred to nitrocellulose membrane (Millipore). Immunoblot analysis was done using primary antibodies. The primary antibodies used were as follows: ERa (sc-544, rabbit polyclonal; Santa Cruz) and BTF3 (sc-10059, goat polyclonal; Santa Cruz). Immunodetection was done using the enhanced chemiluminescence system (Amersham; ref. 42). In vitro Binding Assay Complex Detection of the Interaction between BTFa and ERa by Coimmunoprecipitation For immunoprecipitation in vitro , products were generated in a coupled transcription-translation cell-free proteinsynthesizing reticulocyte lysate with [S]methionine and the expression vector for BTF3a or nonradioactive amino acids and expression vectors for ERa, Tip60-HA, unprogrammed lysate for the empty vector pCMV-HA and immunoprecipitated with monoclonal antibodies to ERa, as well as hemagglutinin antibody as described previously (46). Characterization of the monoclonal antibody to ERa has been previously undertaken (2). The rabbit polyclonal antibody to hemagglutinin-tag raised against an internal region of the influenza hemagglutinin protein (sc-805) also produced the correct-sized proteins on Western blots (Santa Cruz Biotechnology; ref. 45). In detail, cDNA templates for ERa and BTF3a were prepared using a Hybaid kit and resuspended in RNase-free distilled water. An in vitro – coupled transcription and translation kit (T7/T3-TNT; Promega) was used according to the manufacturer’s instructions. After completion of the 90-min reaction, samples were kept on ice. One milliliter of the immunoprecipitation buffer [50 mmol/L Tris-HCl (pH 7.5), 150 mmol/L NaCl, 0.2 mmol/L Na3VO4, 0.5% (w/v) NP40, 1 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L DTT, 25 Ag/mL leupeptin, 25 Ag/mL aprotinin, and 25 Ag/mL pepstatin] were added to each sample, mixed, and incubated on ice for 30 min. Twenty microliters of protein A/G agarose, prewashed thrice in immunoprecipitation buffer, were added to each sample and incubated for an additional 4 h at 4jC with rotation to remove any proteins that had interacted nonspecifically with protein A/G agarose. Protein A/G agarose was removed by centrifugation at 14,000 rpm for 3 min in a bench-top centrifuge. The supernatant containing [S]BTF3a was mixed, in turn with supernatants containing nonradioactive ERa, and the combined supernatants were then incubated with 2.5 Ag of the requisite antibody overnight at 4jC with rotation. Twenty microliters of protein A/G agarose were added to each sample and incubated at 4jC for an additional 60 min. Protein A/G agarose antibody conjugates were recovered by centrifugation at 14,000 rpm for 3 min, resuspended in 1 mL of buffer A [PBS, 0.2% (w/v) Triton X-100, and 350 mmol/L NaCl], and recentrifuged. Samples were resuspended in 1 mL of buffer B [PBS, 0.2% (w/v) Triton X-100], centrifuged, and resuspended in SDS sample buffer. Samples were resolved by electrophoresis on 10% (w/v) polyacrylamide gels at 200 V for 45 min with equal amounts of protein being loaded per lane. The gel was fixed for 30 min in 10% (v/v) propanol, 10% (v/v) acetic acid, dried under vacuum, and exposed to Kodak X-Omat AR X-ray film for 6 to 24 h before developing the film as we have previously described (45). Northern Blots Total RNA was extracted from cells and subjected to Northern blot analysis, as previously described (38, 47). The membranes were hybridized separately and in turn to pBDGAL4 f1, human a-actin, and glyceraldehyde-3-phosphate dehydrogenase probes and subjected to autoradiography. Green et al. Mol Cancer Res 2007;5(11). November 2007 1198 on June 20, 2017. © 2007 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from Transient Transfection of Mammalian Cells The cDNA insert in the pAD-GAL4 f1 plasmid was released by digestion with EcoRI and SalI and ligated with the mammalian expression vector pBK-CMV (Stratagene), digested with the same two enzymes. MCF-7 and HeLa cells were cotransfected with expression and reporter plasmids as previously described (2, 10, 48). Cells were incubated for 24 h without any additions or in the presence of hormones (17h-estradiol, 10 nmol/L, or EGF, 100 ng/mL) and were then harvested and CAT and h-galactosidase activities were assayed as previously described (10). CAT activity results were normalized relative to the h-galactosidase activity, and are expressed as a percentage of the estradiol-induced level, as indicated. Results are the mean F SD of at least three separate experiments. Western Blot Western blots were done on whole-cell extracts of transfected cells, using monoclonal antibodies specific for the amino-terminal or carboxyl-terminal domains of human ERa (Santa Cruz Biotechnology) to confirm that the mutated versions of ERa were each expressed at a level equivalent to that of the wild-type receptor. Statistical Treatment of Results All experiments were done at least thrice. The mean and SD were calculated for each set of results and the significance of the difference between values was assessed by Student’s t test. AcknowledgmentsWe thank S. Ali for hERh0, P. Chambon for HEG0, HE15, HEG19, and HE457,and D.J. Shapiro for ERE.VIT.CAT. We also thank P. Crow and Dr. P. Erwin fortheir technical assistance. References1. El-Tanani MK, Green CD. Oestrogen-induced genes, pLIV-1 and pS2,respond divergently to other steroid hormones in MCF-7 cells. Mol CellEndocrinol 1995;111:75 –81. 2. Katzenellenbogen BS, Choi I, Delage-Mourroux R, et al. Molecularmechanisms of estrogen action: selective ligands and receptor pharmacology.J Steroid Biochem Mol Biol 2000;74:279–85. 3. Tsai MJ, O’Malley BW. Molecular mechanisms of action of steroid/thyroidreceptor superfamily members. Annu Rev Biochem 1994;63:451–86. 4. Tora L, White J, Brou C, Tasset D, Webster N, Scheer E, Chambon P. Thehuman estrogen receptor has two independent nonacidic transcriptional activationfunctions. Cell 1989;59:477 –87. 5. Tzukerman MT, Esty A, Santiso-Mere D, et al. Human estrogen receptortransactivational capacity is determined by both cellular and promoter context andmediated by two functionally distinct intramolecular regions. Mol Endocrinol1994;8:21 – 30. 6. Kumar V, Green S, Stack G, Berry M, Jin JR, Chambon P. Functional domainsof the human estrogen receptor. Cell 1987;51:941–51. 7. Webster NJ, Green S, Jin JR, Chambon P. The hormone-binding domains ofthe estrogen and glucocorticoid receptors contain an inducible transcriptionactivation function. Cell 1988;54:199–207. 8. Bunone G, Briand PA, Miksicek RJ, Picard D. Activation of the unligandedestrogen receptor by EGF involves the MAP kinase pathway and directphosphorylation. EMBO J 1996;15:2174– 83. 9. Ignar-Trowbridge DM, Pimentel M, Parker MG, McLachlan JA, Korach KS.Peptide growth factor cross-talk with the estrogen receptor requires the A/Bdomain and occurs independently of protein kinase C or estradiol. Endocrinology1996;137:1735–44. 10. El-Tanani MK, Green CD. Two separate mechanisms for ligand-independentactivation of the estrogen receptor. Mol Endocrinol 1997;11:928– 37.11. Watanabe M, Yanagisawa J, Kitagawa H, et al. A subfamily of RNA-bindingDEAD-box proteins acts as an estrogen receptor a coactivator through theN-terminal activation domain (AF-1) with an RNA coactivator, SRA. EMBOJ 2001;20:1341 –52. 12. Campbell RA, Bhat-Nakshatri P, Patel NM, et al. Phosphatidylinositol3-kinase/AKT-mediated activation of estrogen receptor a: a new model for anti-estrogen resistance. J Biol Chem 2001;276:9817 –24. 13. Kraichely DM, Sun J, Katzenellenbogen JA, Katzenellenbogen BS.Conformational changes and coactivator recruitment by novel ligands forestrogen receptor-a and estrogen receptor-h: correlations with biologicalcharacter and distinct differences among SRC coactivator family members.Endocrinology 2000;141:3534–45. 14. Deblois G, Giguere V, Su LF, Wang Z, Garabedian MJ. Ligand-independentcoactivation of ERa AF-1 by steroid receptor RNA activator (SRA) via MAPKactivation. J Steroid Biochem Mol Biol 2003;85:123–131. 15. Meyer ME, Gronemeyer H, Turcotte B, Bocquel MT, Tasset D, Chambon P.Steroid hormone receptors compete for factors that mediate their enhancerfunction. Cell 1989;57:433– 42. 16. Tasset D, Tora L, Fromental C, Scheer E, Chambon P. Distinct classes oftranscriptional activating domains function by different mechanisms. Cell 1990;62:1177–87. 17. McKenna NJ, Lanz RB, O’Malley BW. Nuclear receptor coregulators:cellular and molecular biology. Endocr Rev 1999;20:321–44. 18. Toyama T, Iwase H, Yamashita H, et al. p33(ING1b) stimulates thetranscriptional activity of the estrogen receptor a via its activation function (AF) 2domain. J Steroid Biochem Mol Biol 2003;87:57– 63. 19. Endoh H, Maruyama K, Masuhiro Y, et al. Purification and identification ofp68 RNA helicase acting as a transcriptional coactivator specific for theactivation function 1 of human estrogen receptor a. Mol Cell Biol 1999;19:5363– 72. 20. Freedman LP. Multimeric coactivator complexes for steroid/nuclear recep-tors. Trends Endocrinol Metab 1999;10:403–7. 21. Lanz RB, McKenna NJ, Onate SA, et al. A steroid receptor coactivator,SRA, functions as an RNA and is present in an SRC-1 complex. Cell 1999;97:17 –27. 22. Shibata H, Spencer TE, Onate SA, et al. Role of co-activators and co-repressors in the mechanism of steroid/thyroid receptor action. Recent Prog HormRes 1997;52:141 –65. 23. Torchia J, Glass C, Rosenfeld MG. Co-activators and co-repressors inthe integration of transcriptional responses. Curr Opin Cell Biol 1998;10:373 –83. 24. Webb P, Nguyen P, Shinsako J, et al. Estrogen receptor activationfunction 1 works by binding p160 coactivator proteins. Mol Endocrinol 1998;12:1605–18. 25. Onate SA, Boonyaratanakornkit V, Spencer TE, et al. The steroid receptorcoactivator-1 contains multiple receptor interacting and activation domains thatcooperatively enhance the activation function 1 (AF1) and AF2 domains ofsteroid receptors. J Biol Chem 1998;273:12101–8. 26. Kato S, Endoh H, Masuhiro Y, et al. Activation of the estrogen receptorthrough phosphorylation by mitogen-activated protein kinase. Science 1995;270:1491– 4. 27. Ali S, Metzger D, Bornert JM, Chambon P. Modulation of transcriptionalactivation by ligand-dependent phosphorylation of the human oestrogen receptorA/B region. EMBO J 1993;12:1153–60. 28. Chien CT, Bartel PL, Sternglanz R, Fields S. The two-hybrid system: amethod to identify and clone genes for proteins that interact with a protein ofinterest. Proc Natl Acad Sci U S A 1991;88:9578 –82. 29. Zheng XM, Black D, Chambon P, Egly JM. Sequencing and expression ofcomplementary DNA for the general transcription factor BTF3. Nature 1990;344:556 –9. 30. El-Tanani MK, Fernig DG, Barraclough R, Green C, Rudland P. Differentialmodulation of transcriptional activity of estrogen receptors by direct protein-protein interactions with the T cell factor family of transcription factors. J BiolChem 2001;276:41675– 82. 31. Zheng XM, Moncollin V, Egly JM, Chambon P. A general transcriptionfactor forms a stable complex with RNA polymerase B (II). Cell 1987;50:361 –8. 32. Brady ME, Ozanne DM, Gaughan L, et al. Tip60 is a nuclear hormonereceptor coactivator. J Biol Chem 1999;274:17599–604. 33. Chang TC, Nardulli AM, Lew D, Shapiro DJ. The role of estrogen responseelements in expression of the Xenopus laevis vitellogenin B1 gene. MolEndocrinol 1992;6:346– 54.BTF3 Interacts with Human Estrogen Receptor a Mol Cancer Res 2007;5(11). November 20071199 on June 20, 2017. © 2007 American Association for Cancer Research.mcr.aacrjournals.orgDownloaded from 34. Tora L, Mullick A, Metzger D, Ponglikitmongkol M, Park I, Chambon P. Thecloned human oestrogen receptor contains a mutation which alters its hormonebinding properties. EMBO J 1989;8:1981–6. 35. Segraves WA. Something old, some things new: the steroid receptorsuperfamily in Drosophila. Cell 1991;67:225– 8. 36. Danielian PS, White R, Lees JA, Parker MG. Identification of a conservedregion required for hormone dependent transcriptional activation by steroidhormone receptors. EMBO J 1992;11:1025–33. 37. Kraus WL, McInerney EM, Katzenellenbogen BS. Ligand-dependent,transcriptionally productive association of the aminoand carboxyl-terminalregions of a steroid hormone nuclear receptor. Proc Natl Acad Sci U S A 1995;92:12314 –18. 38. El-Tanani MK, Platt-Higgins A, Rudland PS, Campbell FC. Ets gene PEA3cooperates with h-catenin-Lef-1 and c-Jun in regulation of osteopontintranscription. J Biol Chem 2004;279:20794–806. 39. Deng JM, Behringer RR. An insertional mutation in the BTF3 transcriptionfactor gene leads to an early postimplantation lethality in mice. Transgenic Res1995;4:264 –9. 40. Hu GZ, Ronne H. Yeast BTF3 protein is encoded by duplicated genes andinhibits the expression of some genes in vivo . Nucleic Acids Res 1994;22:2740–3. 41. Cavailles V, Dauvois S, L’Horset F, et al. Nuclear factor RIP140modulates transcriptional activation by the estrogen receptor. EMBO J 1995;14:3741– 51. 42. El-Tanani MK, Barraclough R, Wilkinson MC, Rudland PS. Regulatoryregion of metastasis-inducing DNA is the binding site for T cell factor-4.Oncogene 2001;20:1793–7. 43. Russo GL, Tosto M, Mupo A, Castellano I, Cuomo A, Tosti E. Biochemicaland functional characterization of protein kinase CK2 in ascidian Cionaintestinalis oocytes at fertilization. Cloning and sequence analysis of cDNA fora and h subunits. J Biol Chem 2004;279:33012 –23. 44. Olesen JT, Fikes JD, Guarente L. The Schizosaccharomyces pombe homologof Saccharomyces cerevisiae HAP2 reveals selective and stringent conservationof the small essential core protein domain. Mol Cell Biol 1991;11:611 –9. 45. Pace P, Taylor J, Suntharalingam S, Coombes RC, Ali S. Human estrogenreceptor h binds DNA in a manner similar to and dimerizes with estrogen receptora. J Biol Chem 1997;272:25832–8. 46. El-Tanani MK, Campbell FC, Crowe P, et al. BRCA1 suppressesosteopontin-mediated breast cancer. J Biol Chem 2006;281:26587 –601. 47. El-Tanani MK, Green CD. Insulin/IGF-1 modulation of the expression of twoestrogen-induced genes in MCF-7 cells. Mol Cell Endocrinol 1996;121:29–35. 48. El-Tanani M, Barraclough R, Wilkinson MC, Rudland PS. Metastasis-inducing DNA regulates the expression of the osteopontin gene by binding thetranscription factor Tcf-4. Cancer Res 2001;61:5619 –29.Green et al. Mol Cancer Res 2007;5(11). November 20071200 on June 20, 2017. © 2007 American Association for Cancer Research.mcr.aacrjournals.orgDownloaded from 2007;5:1191-1200.Mol Cancer ResChris D. Green, Paul D. Thompson, Patrick G. Johnston, et al.αHuman Estrogen Receptor-Terminal Domain of2Transcription Factor 3, and the NHInteraction between Transcription Factor, Basal Updated versionhttp://mcr.aacrjournals.org/content/5/11/1191Access the most recent version of this article at: Cited articleshttp://mcr.aacrjournals.org/content/5/11/1191.full.html#ref-list-1This article cites 48 articles, 16 of which you can access for free at: Citing articles/content/5/11/1191.full.html#related-urlsThis article has been cited by 1 HighWire-hosted articles. Access the articles at: E-mail alertsrelated to this article or journal.Sign up to receive free email-alerts SubscriptionsReprints [email protected] atTo order reprints of this article or to subscribe to the journal, contact the AACR Publications [email protected] atTo request permission to re-use all or part of this article, contact the AACR Publications on June 20, 2017. © 2007 American Association for Cancer Research.mcr.aacrjournals.orgDownloaded from